The No.1 Website for Pro Audio
 Search This Thread  Search This Forum  Search Reviews  Search Gear Database  Search Gear for sale  Search Gearslutz Go Advanced
Dave Rats concept of incoherent speakers Synchronisers/Clock Generators
Old 1 week ago
  #1
Lives for gear
 
Christof's Avatar
Dave Rats concept of non-coherent speakers

Today I've been simulating different speaker-setups with EASE Focus (see pic attached). Both setups have a CSA-system in the center, but I was interested in the difference between flying the tops as a center cluster or left-right as usual. The left-right-setup would give me better GbF, because the distance between the center of the stage and the speakers hanging at the side is much greater. I would prefer the L-R-setup, but interference will causing heavy combfiltering with varying constructive and destructive zones. Hanging both tops center would reduce interference subtantially.

While thinking about that I remembered a video I saw some time ago on Dave Rats Youtube-channel:


I really like how Dave Rat demonstrates thing in a very practical way - and it's an interesting concept. But that got me thinking: How can this be practically applied to the usual "small stage" gig (in our case it's a 8x6m open-air-stage with a bigband, some choirs...)?

Comb-filtering and interference can be greatly reduced, if:
a) both signals are not coherent
b) the signals level on one side is substantially lower than on the other side (reminding me of the 3:1-rule for setting up mics)

If you got a stereo-signal from stereo drum-overheads or a electric piano with a stereo-output (being incoherent hopefully) that seems to be easy.

BUT: What about kick/snare, bass-guitars and lead-vocals being mono-sources that need to be heard upfront?
Dave Rat seems to use different mics for kick and snare and pans those left-right, but ìn many cases one would not have enough channels or mics to do this. And of course I can't imagine a vocalist singing into two microphones.

What do you think about that concept? Did you integrate some of those ideas into your practical work and if you did could you explain some more?

EDIT: The frequency-response-graphs are showing 3 different listening positions: The blue one being a position of a person standing some meters in front of the stage, the others being positions on stage (front center, mid center).
Attached Thumbnails
Dave Rats concept of incoherent speakers-setup-comparison.jpg  

Last edited by Christof; 1 week ago at 10:23 AM..
Old 1 week ago
  #2
Lives for gear
 
Dutchy15's Avatar
Welcome to the world of PA system design! It seems you have already met our biggest and most important friend in this field; the compromise. No design is ever without compromise, the art is in choosing the best compromise for the situation. A left-right will, in any situation, cause an interference pattern. The severeness of the interference will depend on the frequency of the signal (try using higher frequencies in EASE). It is a much bigger issue for subs than it is for tops. The interference will still occur but unless you'd like to live with a mono PA you'll have to go stereo. (mono vs stereo in live sound has been discussed more than elaborately on this and other forum websites, please don't start again here)

The added benefit of sticking with traditional L/R is the increased coverage of the system, a more even SPL can be provided across the audience area as 6dB/doubling of distance always applies. Center subs create an alignment compromise in exchange for great coverage and virtually full control over interference (so you can use it in your advantage, i.e. arraying).

If there's no time for elaborate research on your side I'd recommend sticking with a full L/R setup where each side features 3 subs and a top (judging from your EASE). The subs can be in CSA, end-fired or just regular but it'll make for much easier time alignment compared to center subs.

If you're really interested in this stuff I'd recommend Bob McCarthy's "green bible"; buy once read (at least) twice. Also, Mapp XT is great free software for messing around with sub arrays and setups (not sure on EASE's capabilities in that department). D&B and L'Acoustics also make great simulators / calculators.


Dutchy
Old 1 week ago
  #3
Lives for gear
 
Christof's Avatar
Quote:
Originally Posted by Dutchy15 View Post
Welcome to the world of PA system design! It seems you have already met our biggest and most important friend in this field; the compromise.
That's a nice way to say - finding the best compromise for a given applicatin is the goal we all pursue.

Thanks for pointing out that time-aligning center subs with L-R-tops is a problem / only valid for one position. So could you're recommendation be read as: Go with L-R-subs, if you choose L-R-tops or go with center tops if you use centered subs? Seems to make sense...

In our case there's no big difference in coverage of both setups. It's a small open air stage (8x6m oder 10x6m) and the space for the audience is limited (especially on the sides). So flying speakers L-R and turning them inwards would give a similar coverage compared to hanging them center and turning them outwards (still being stereo, but probably with a reduced perceived width).

I don't want to discuss mono/stereo here, still it seems to be common ground that you won't do extreme things on a live stage (like one instrument panned hard on one side) and of course the most important signals are expected to be coming from the center and can be viewed as being "mono" (lead vocals, bass, kick/snare...). So from this perspective I think one can’t ignore that coherent signals coming from two speakers will always cause heavy interference/comb-filtering if you’re not standing at a “perfect position“ (with more obvious effects at longer wavelengths - but not limited to subs).

We’ve been choosing a centerd CSA-setup because we wanted to avoid the heavy lobing of L-R-subs we found to be disturbing (placing a CSA-system each on both sides would not work as good). A litte piece of applied acoustics that worked really great in the past! EASE Focus is a good tool to experiment and optimize (will have a look at Mapp XT too).

Dave Rats concept of incoherent signals on both sides sounds attractive because it would make a big difference in this area - if it’s practically feasible - as interference could be greatly reduced (and with stereo-signals everyone's already doing that without thinking about it). But with signals it makes things complicated and with others - like lead vocals - this doesn’t seem to be realistic at all – maybe that’s where a L-C-R-system would make sense.
Old 1 week ago
  #4
Lives for gear
 

Thank you Dutchy 15 for a coherent contribution to a discussion pursuant to a journey that entails lots of pot holes. I frequently do one mic guitar/vocal performances that are obviously well suited to mono delivery. My KV2 ES system, if properly placed, delivers L-R bass very well however if not arranged pursuant to factory recommendations, problems can and do occur. The precise directivity of the ES point source system is very important to this end but occasionally when I "sit in" on some less capable systems the sonic degradation can make my single mic performance difficult. My friend Cliff Miller, owner of SE systems, was the main SR provider for Alison Krauss and he used very small L-R panning of every thing but lead vocals.These were not simple L-R systems: several trailer loads of gear was in place when the sound check would start. However the sonic quality of those world wide concert tours are legendary in the acoustic Americana world and the extremely judicious L-R panning was a critical element to that end.
Hugh
Old 1 week ago
  #5
Lives for gear
 
Dutchy15's Avatar
I haven't tried Dave Rat's incoherent signal theory myself but I could give you a few arguments as to why it is not as easy as it seems. The main one is that low frequencies tend to be very sinewave-like (bass guitars, kick drums, synths, etc). Even when using two different mics on a drum or a mic and a DI on a bass guitar the low end will still be much too coherent to not interfere acoustically when the two channels are panned hard L/R. It is the same source after all. One might argue that phase differences (different mics / positions) and arrival times from source to capture (mic/DI, positions) will make the signals less coherent but a sine is a sine. Also, since sine waves (and triangle, square, etc) are very periodic the interference will also occur if one of the two signals is delayed by one or several periods. A sine is a sine.

Dave uses (pink) noise in his experiment, which is fully random and thus non-periodic. I'd be interested in the outcome of the same experiment using some small headphone drivers (representative frequency spectrum reproduction) and real life source material.

Again, my advice would be to stick with a full L/R setup for both subs and tops. Center subs are possible but make for a more difficult time alignment between subs and tops since it is only valid in one position / area. I wouldn't worry about the theoretical interference and just have a listen

Theory and simulations are great for rainy Sundays like this one here in the Netherlands but there's (quite a few) reasons that PA's have been L/R tops for the last couple of decades. Sub setups have been all over the place although some of the best ones I have heard were either L/R or full stage width horizontal arrays. As SamC would preach; KISS.

By the way; what PA system are you using and what processor & amps? Stuff like linear-phase EQ and all-pass filters makes tuning systems so much easier.


Dutchy
Old 1 week ago
  #6
Lives for gear
 
Christof's Avatar
Quote:
Originally Posted by Dutchy15 View Post
I haven't tried Dave Rat's incoherent signal theory myself but I could give you a few arguments as to why it is not as easy as it seems. The main one is that low frequencies tend to be very sinewave-like (bass guitars, kick drums, synths, etc). Even when using two different mics on a drum or a mic and a DI on a bass guitar the low end will still be much too coherent to not interfere acoustically when the two channels are panned hard L/R. (...)

Dave uses (pink) noise in his experiment, which is fully random and thus non-periodic. I'd be interested in the outcome of the same experiment using some small headphone drivers (representative frequency spectrum reproduction) and real life source material.
Thanks for yor explanations, that sounds perfectly plausible. It's still an interesting theory, but not practical for usual stage-setups.

Quote:
Originally Posted by Dutchy15 View Post
Again, my advice would be to stick with a full L/R setup for both subs and tops. Center subs are possible but make for a more difficult time alignment between subs and tops since it is only valid in one position / area. I wouldn't worry about the theoretical interference and just have a listen
I've been thinking about CSA or other sub-configurations since I heard the heavy interference of the subs while walking sidewards in front of the stage. At the mixing-desk the bass was fat and just a few meters aside it wasn't there.
As a result I had been using a CSA-setup of 2 rows with 3 subs each in front of the stage. The bass dispersion had been so much better and on stage - just behind the subs - the bass had been at least 20dB down. Really liked that, so it's the listening-experience that makes me want to setup the subs like that again. But as you mentioned, time aligning with L-R-speakers is not really possible and so that's why I'm thinking about hanging the tops center. Time-alignment should be much easier this way.

Quote:
Originally Posted by Dutchy15 View Post
Theory and simulations are great for rainy Sundays like this one here in the Netherlands but there's (quite a few) reasons that PA's have been L/R tops for the last couple of decades.
How did you know? Living in north-germany we had heavy rain and stormy wind all day... and I always like thinking aside from the usual ways.

Quote:
Originally Posted by Dutchy15 View Post
By the way; what PA system are you using and what processor & amps? Stuff like linear-phase EQ and all-pass filters makes tuning systems so much easier.
As I mentioned before it's a small stage (8x6m or 10x6m) and the audience is standing within 10m from the stage. We bought the smallest and lightest system with good sound that's able to deliver the SPL needed (in germany the regulation of DIN 15905-5 is limiting the SPL allowed to 95dB averaged over half an hour at the loudest position the audience can get to - we're using a special software, mic and calibration-device to measure and document the SPL).

The tops are HK Audio Linear 5 112FA (12"-woofer, asymetrical horn, used with tilt brackets for flying) and our tiny subs are "the Box Pro Achat 115" - not expensive, but giving us different options to set them up - and they sound good. Max. SPL is limited, but for our purpose it's just what we need.

The subs are driven with a Dynacord SL-2400 (same as EV Q-1212 and capable of driving the subs at 2,6 Ohm), the processing is done with an EV DX-38 or just through the matrix-channels of an X32-desk (no linear-phase-EQs) - so rather "budget" equipment, but still reasonable.
Old 1 week ago
  #7
Lives for gear
 
Wyllys's Avatar
 

I'd prefer differentiated or non-coherent rather than incoherent...
Old 1 week ago
  #8
Lives for gear
 

i'd call his concept 'incoherent', not the speaker setup;
same applies to his mix (at least when i heard him mix the chili peppers...)
Old 1 week ago
  #9
Lives for gear
 
Christof's Avatar
Yes of course both of you are right, that's a better way to say it.

After we made that clear - what do you think about his concept?
Or regarding my thoughts about setting up the speakers: How would you deal with the tops, if using a centered CSA-system?

Last edited by Christof; 1 week ago at 01:54 PM.. Reason: typo
Old 1 week ago
  #10
Lives for gear
 

i'm a huge fan of l/C/r systems (with a capital C!) but they have their own bunch of issues - there is no setup that would combine all concepts and would work in all venues.

my points of critique to dave's approach is that if you put two mics on a single source and pan signals to opposite sides, you need to make sure the levels are pretty much the same (or your mix will tip to one side): not always easy to confirm if foh is off center. also, were now talking about dual mono mixing: difficult to control on a stereo or l/c/r system (as you cannot switch off one leg during a show just to control). and it takes quite a bit more gear if you want to practice this approach to full extent, something that simply cannot be done on small productions (gear, costs, time, manpower become too expensive).
and then, one would need to consider how much can be gained: system design/setup/overlap have a large influence on results - but even if if there's something to be gained, what will you do with mono frontfills and delay lines? drive everything off a large matrix and decide what source will be best for each mix? how do these mixes combine with the less correlated mix in the area of overlap?
maybe if you're headlining and can push through all requests, have enough time in production rehearsals, why not give it shot on some sources which you think will profit, yet although i never tried this appoach to full extent, i'm almost convinced it ain't worth the pain...

Last edited by deedeeyeah; 1 week ago at 11:36 PM.. Reason: typo
Old 1 week ago
  #11
Lives for gear
Quote:
Originally Posted by Christof View Post
Yes of course both of you are right, that's a better way to say it.

After we made that clear - what do you think about his concept?
Or regarding my thoughts about setting up the speakers: How would you deal with the tops, if using a centered CSA-system?
Even if you have your mains on a stick or flown above the subs on each side, you will only get perfect addition on the point you use to align the phase between main and subs in the room. One meter to the side and you won't be perfectly in phase anymore.
So if you center your subs or make a bass array, imagine your subs as one system and take the middle of the bass array as your source for aligning (and in case of a sub array add the delay to the whole array).

Also, as phase alignment of subs to tops at the distance from the pa where it makes sense to be in phase to gain some db of sub, is a pain in the ass (most likely coherence will be bad). So measure your delay between main and sub with the main speaker standing directly on top of the sub and align there (with a distance of the mic to the speakers of about a meter) and note that delay.
Than, with your PA hung and stacked as you liked, meassure the distance between main and sub with a laser meter and add that to the delay you measured before.

Last edited by tenderboy; 1 week ago at 11:45 PM..
Old 1 week ago
  #13
Lives for gear
And here is a nice tool:
Merlijn van Veen - sub align
Old 1 week ago
  #14
Lives for gear
 

I'll admit that I've only used the standard EASE package, not the Focus version . . . but a quick observation of your plots:
- From the sharpness of the comb-filter artifacts, it looks like you're running these simulations at 1/3rd octave bandwidth
- Given the minimal directivity shown, you've chosen a frequency range that's well below the crossover point of the tops
Because of this, your illustration gives very little useful information about how these arrangements will actually perform.

The most important thing in speaker layout is, quite simply,to put the sound on the people, not on the walls, and not into the microphones. If you're not accomplishing this, then all of the musing about comb-filtering and cross-channel coherence makes absolutely no difference.

Shoebox-shaped rooms can be tough to get right for a number of reasons, but if I read the general scale of your room correctly, my instinct is that a L/R setup will better leverage the horns' geometry to keep sound off the side walls, and avoid the tendency for them to couple together and cause a buildup of lower-midrange/upper-bass energy directly below, as tends to happen in a center cluster. I would look at in simulation with two sets of runs . . . the first with centered around 3KHz, to give a clear idea of what the horns are doing, and then another set at perhaps 250Hz to see the effects in the range where the speakers have virtually no pattern control. If you use i.e. three-octave bandwidths, you'll lose that dramatic illustration of the comb-filtering effects, and end up with something that's more in line with what you'll actually experience subjectively.

Also don't forget that when you combine several speakers in a cluster . . . in the real world they don't magically end up at a single point acoustically, like you can do in simulation. The proper way to examine this with EASE is to first build a cluster in Speaker Base/Speaker Lab, with the exact distances between their acoustic centers that they'll have when rigged together, simulate the polar response of the whole array, then drop this into your project. You'll find that there are very few 12"-and-a-horn tops on the market that are at all usable when combined in a cluster . . . usually the aggregate directivity is horribly lumpy, even if it's nice and smooth for a single speaker.

So it's quite likely that by putting your existing speakers together in a center array, all in an effort to reduce comb-filtering effects of L/R systems . . . you'll simply be turning them into a very poor-performing cluster. Not a good tradeoff.
Old 1 week ago
  #15
Lives for gear
 
Christof's Avatar
@tenderboy: Will have a closer look at it tomorrow.

@kirkus: Thanks for your helpful contribution! Some explanations:
The stage is not indoors, but open air. There are some buildings on the sides not too far away, but because of the way the stage is setup relative to the walls reflections will be no major concern. In front and behind the stage there are no buildings.

My goal is to make good use of height/tilt of the tops to focus on the area about 10-12m in front of the stage (because there are other activities further away that don't want loud sound) and to get a good front-to-back-ratio in this area for an even coverage.

EASE Focus shows that the backward rejection of the tops is getting substantially effective from about 1k upwards, the crossover to the tweeter is at 1.65kHz. My concern about the comb-filtering initially arised because of how it might affect the microphones on stage at frequencies where the fundamentals of voices and many instruments are. I need to maximize GbF, because there'll be a lot of open mics on stage. As keyboards and bass/guitars could/will be placed L-R at the front of the stage (not sensitive for feedback) the L-R-setup would probably work best giving me some more distance to the mics (setting up the mics in a way to get best backward rejection is mandatory of course).

The graphics show the 160Hz 1/3rd octave band (where the tops take over, crossover had been 125Hz - possible with those subs and takes some stress away from the 12"-tops). How to deal with time-alignment and crossover between the subs centered and the tops L-R ist still giving me some thoughts...

I'm probably turning away from the center cluster. But shouldn't EASE Focus be capable of simulating such a cluster using the gll-data of the speakers? By the way: Using the gll-viewer I once found out, that the quality of the data differs considerably between different speakers. Some of them are very generously smoothed, others are very detailed; the gll-file of the tops we use is one of the detailed ones.
Old 6 days ago
  #16
Lives for gear
 
Christof's Avatar
So after thinking about Dave Rats concept and some interesting theoretical things, I rather did some more experimenting with EASE Focus and it looks like I found a good and practical solution for our purpose:

Subs as CSA or EFA-setup centered in front of the stage (pics show 80Hz-behaviour)
  • Setup like a "Y" with one stack centered about 1m behing two subs in front building the "wings", spaced about 1.4m
  • 2 subs each stacked on one another (6 altogether)
  • Delay-time 3ms (CSA/EFA depending on which row is delayed in relation to the other)

EASE Focus shows a broad dispersion of both types and a good backward rejection (with the CSA being better in this regard). Coverage in most 1/3-octave-bands is about the same as coverage of the tops (L-R-setup, mounted to the front truss of the stage). Switching between the CSA/EFA-type is just a matter of changing the delay-times of the subs, so it will be easy to compare and decide.
Attached Thumbnails
Dave Rats concept of incoherent speakers-setup-csa.jpg   Dave Rats concept of incoherent speakers-setup-efa.jpg  
Top Mentioned Products
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Similar Threads
Thread
Thread Starter / Forum
Replies
bluntforce / So much gear, so little time
0
kristoffer / So much gear, so little time
10
groovenagl / Product Alerts older than 2 months
20
d1rtynyc / The Moan Zone
16

Forum Jump
Forum Jump